The Diabetes Educator

A Family-Based Diabetes Intervention for Hispanic Adults and Their Family Members Jie Hu, Debra C. Wallace, Thomas P. McCoy and Karen A. Amirehsani The Diabetes Educator 2014 40: 48 originally published online 18 November 2013 DOI: 10.1177/0145721713512682 The online version of this article can be found at:

Published by:

On behalf of:

American Association of Diabetes Educators

Additional services and information for The Diabetes Educator can be found at: Email Alerts: Subscriptions: Reprints: Permissions:

>> Version of Record - Jan 17, 2014 OnlineFirst Version of Record - Nov 18, 2013 What is This?

Downloaded from at Nat. Taichung Univ. of Sci. & Tech. on April 21, 2014



TDEXXX10.1177/0145721713512682A Family-Based Diabetes InterventionHu et al

The Diabetes EDUCATOR 48

A Family-Based Diabetes Intervention for Hispanic Adults and Their Family Members Aims

Jie Hu, PhD, RN Debra C. Wallace, PhD, RN, FAAN Thomas P. McCoy, MS, PStat Karen A. Amirehsani, PhD, FNP-BC From the University of North Carolina at Greensboro, Greensboro, North Carolina. Correspondence to Jie Hu, PhD, RN, School of Nursing, University of North Carolina at Greensboro, P.O. Box 26170, Greensboro, NC 27402-6170 ([email protected]). Acknowledgments: The project described was supported by grant P20MD002289 from the National Institute for Minority Health and Health Disparities. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute for Minority Health and Health Disparities or the National Institutes of Health. Appreciation is extended to our community partners for assisting with the study logistics. DOI: 10.1177/0145721713512682 © 2013 The Author(s)

The purpose of this quasi-experimental, 1-group longitudinal study is to examine the effects of a family-based intervention program on diabetes self-management behaviors, A1C, other biomarkers, psychosocial factors, and health-related quality of life in Hispanics with diabetes.

Methods Adult patients with diabetes (n = 36) and family members (n = 37) were recruited from a community clinic in rural central North Carolina. Patients and family members attended an 8-week culturally tailored diabetes educational program taught in Spanish. Data were collected pre- and post-intervention for both patients and family members, with an additional data collection for patients 1 month post-intervention.

Results Most patients and family members were female, and almost all were immigrants. A1C decreased by 4.9% on average among patients from pre-intervention to 1 month post-intervention. Patients showed significant improvements in systolic blood pressure, diabetes self-efficacy, diabetes knowledge, and physical and mental components of health-related quality of life. Higher levels of intake of healthy foods and performance of blood glucose tests and foot inspections were reported. Family members significantly lowered body mass index and Volume 40, Number 1, January/February 2014 Downloaded from at Nat. Taichung Univ. of Sci. & Tech. on April 21, 2014

A Family-Based Diabetes Intervention 49

improved diabetes knowledge from pre-intervention to immediately post-intervention. No significant changes in levels of physical activity were found among patients with diabetes or family members.

Conclusions Findings suggest that including family members in educational interventions may provide emotional and psychological support to patients with diabetes, help to develop healthy family behaviors, and promote diabetes self-management.


ispanics and Latino Americans are 66% more likely to develop diabetes than nonHispanic whites1 and suffer disproportionately from diabetes-related complications, including hypertension, heart disease, stroke, kidney disease, blindness, nervous system disease, and amputation.2 Hispanics have significantly higher rates of hospitalization for uncontrolled diabetes and complications than do whites and are 1.5 times more likely to die from diabetes.3 Diabetes self-management is key to achieving glycemic control and improving health outcomes.4 There is strong evidence that following a healthy diet and exercise program, taking diabetes medications, and monitoring glucose can improve diabetes outcomes.1 However, Hispanics with type 2 diabetes (T2DM) show poorer self-management of the disease than non-Hispanic whites; only 36.8% of Mexican Americans with diabetes have their hemoglobin A1C under control, compared with 60% of non-Hispanic whites.3 Many Hispanics fail to adhere to the diabetes selfmanagement recommendations of the American Diabetes Association (ADA).5 Hispanics are 2.09 times less likely than whites to perform adequate physical activity6 and 36% less likely to perform self-monitoring of glucose.5,7 Clearly, interventions are needed to enhance diabetes-self management in Hispanic Americans. Social and cultural influences play a large role for Hispanics in dealing with diabetes.8 In particular, familismo is an important value: “The family comes first.” Thus, focusing on family involvement and family centeredness is an important aspect of interventions for Hispanics with diabetes.9,10 Hispanics face many challenges in making the

lifestyle changes required for effective and sustained diabetes self-care.11,12 For these patients, social support, including peer and family involvement, has been related to improved healthy eating and improved physical activity,13 improved low-density lipoprotein (LDL) and total cholesterol,14 better glycemic control as measured by A1C,15 improved knowledge,16 improved self-efficacy17 and better self-management.18 Social cognitive theory, which suggests that behavior change depends on the interaction of personal factors, the environment, the situation, and behavior, guided the intervention. Environment includes both physical (barriers to diabetes self-care) and social (family members) factors. Situations refer to an individual’s perception of the environment and how this influences his or her behavior, including perception of support from family members. Self-efficacy refers to an individual’s confidence to perform diabetes self-care activities. Behavior includes diabetes self-management. Our intervention focused on increasing diabetes knowledge, overcoming barriers to self-management, and fostering behavioral changes through family support and development of selfefficacy. Therefore, we expected that after participants received the family-based intervention, they would demonstrate the following: 1. Improvements in self-management of diabetes, including physical activity, diet, and self-monitoring of blood glucose 2. Improvements in metabolic measures of body mass index (BMI), A1C, total cholesterol/high-density lipoprotein (HDL) ratio, and blood pressure 3. Increases in diabetes knowledge and in scores on perceived self-efficacy, family support, and health-related quality of life (physical and mental health)

Methods Design

A quasi-experimental, 1-group longitudinal design was used as a pilot study to examine the effects of the family-based intervention program. For participants with diabetes, data were collected at baseline, post-intervention, and at 1-month follow-up. For family members, data were collected at baseline and post-intervention. Sample and Setting

Participants with T2DM and at least 1 family member were recruited from a community clinic for the uninsured

Hu et al Downloaded from at Nat. Taichung Univ. of Sci. & Tech. on April 21, 2014

The Diabetes EDUCATOR 50

that serves a large population of Hispanics in rural central North Carolina. Criteria for inclusion for patients with diabetes included (a) self-identity as Hispanic, (b) age 18 years or older, (c) self-report of a medical diagnosis of T2DM, and (d) an adult family member willing to participate. Inclusion criteria for family members were (a) residence in the patient’s household and (b) age 18 years or older. Participants and family members had to be able to speak either Spanish or English. Those who were pregnant, were diagnosed with type 1 diabetes, reported prior (past year) or current participation in other diabetes selfmanagement intervention programs, or were cognitively impaired were excluded. Participants were recruited through flyers distributed by clinic staff and through waiting room conversations with a bilingual and bicultural interpreter. Potential participants met a research assistant or the first author in a private room or made an appointment for a home visit/ family session with the interpreter and research assistant/ first author for consent and enrollment. Incentives were store gift cards. The university institutional review board approved the study. Intervention

The intervention was based on social cognitive theory, which focuses on the influence of the social environment (family members) in health behavior changes to overcome barriers to diabetes self-management and enhance selfefficacy. The intervention consisted of 2 family sessions (pre- and post-intervention for data collections) and 8 weekly educational group sessions for participants and family members conducted by a nurse practitioner/trained educator bilingual in Spanish. For the 2 family sessions, the family unit, including multiple members, was invited. The first family session explained the purpose of the study, the format of the intervention, and requirements of participants. Informed consent was obtained from the participant and family member, and baseline data were collected. After that, each participant was asked to bring at least 1 family member to the 8 group intervention meetings. The last family session included post-intervention data collection and discussions with family members about diabetes self-management. The 8 weekly interactive modules were modified from a family-based diabetes program19 reflecting the National Standards for Diabetes Self-Management Education and Support20 and the National Diabetes Education Program

( The modules also included educational materials, strategies, activities, and information from the ADA, the National Center for Health Statistics (Centers for Disease Control and Prevention), and US Department of Health and Human Services. Approaches to success through family support were provided throughout the intervention. The 8 modules were titled Introduction to Diabetes; Exercise and Food; Eating Healthy; Blood Sugar Levels and Glucometers; Diabetes Medications; Taking Care of Your Body; Coping Strategies, Problem Solving, and Action Plans; and Summary & Action Plan for You and Your Family. A group discussion with open-ended questions on family support was facilitated at the end of the first and last group meetings. All of the 8 modules were tailored to low-literacy needs and integrated cultural beliefs and values. Bilingual nurses and interpreters taught about dietary change, using modified ethnic foods and recipes and culturally relevant activities.21 Picture illustrations, seminar discussions, educational flipcharts and games, videotapes, visual aids (pictorial log sheets, pictorial food books), demonstrations, and self-monitoring demonstrations were used. The last intervention session included a celebration of completion of the program for both participants and family members with a certificate and food. Instruments

Demographic forms included family history, health history, socioeconomic information, and the number and frequency of family members attending the home visits and group meetings. Hemoglobin A1C was tested with a Bayer A1C NOW kit (Bayer Healthcare, Sunnyvale, CA) using finger stick blood taken by a registered nurse; A1C greater than 7.0% (53 mmol/mol) indicated poor glycemic control.4 Fasting glucose and lipid profiles, including total cholesterol, HDL, LDL, and triglycerides (TG), were obtained through capillary finger stick. A Cholestech LDX machine (Alere, Inc., Waltham, MA) was used for glucose and lipid profiles. The accuracy and precision of the Cholestech profiles are comparable to those obtained by reference methods used routinely in clinical diagnostic laboratories.22 The Cholestech machine was calibrated at each session to ensure the accuracy of readings. Height and weight were measured using standard procedures,23 and BMI was calculated. A participant was categorized as overweight if BMI was 25 to 29.9 kg/m2 and obese if BMI was 30 kg/m2 or more.24 Waist circumference was

Volume 40, Number 1, January/February 2014 Downloaded from at Nat. Taichung Univ. of Sci. & Tech. on April 21, 2014

A Family-Based Diabetes Intervention 51

measured using a Gantt tape measure. Measurements were recorded to the nearest 0.1 cm.24 Participants were categorized as high risk if waist circumference was greater than 40 inches (102 cm) for men and greater than 35 inches (88 cm) for women.24 Blood pressure was measured using a standardized blood pressure protocol based on the guidelines of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure.25 Participants were considered hypertensive if they used antihypertensive medications or had blood pressure 130/80 mm Hg or greater. Physical activity was measured using the Short International Physical Activity Questionnaire (IPAQ) form, Last 7 Days Recall, a 9-item scale. The short IPAQ provides information on time spent walking, participating in vigorous and moderate-intensity activities, and participating in sedentary activities in the last 7 days. Energy expenditure was calculated using estimated metabolic equivalent task (MET)-minutes per week. Walking MET-minutes per week were calculated as 3.3 × walking minutes × walking days; moderate activity MET-minutes per week were calculated as 4.0 × moderate-intensity activity minutes × moderate activity days; and vigorous MET-minutes per week were calculated as 8.0 × vigorous activity minutes × vigorous activity days. Estimates of MET-minutes per week for total physical activity were then calculated as the sum of walking, moderate, and vigorous MET-minutes per week. Kilocalories per week were estimated as total physical activity MET-minutes per week × (weight in kilograms / 60). Total physical activity MET-minutes per week and kilocalories per week were subsequently analyzed. Reliability and validity for the IPAQ have been established in 12 countries26 and for patients with T2DM.27 Diet was measured using items from the Behavioral Risk Factor Surveillance System (BRFSS) survey (6 items),28 which include daily intake of fruits and vegetables. Each item was measured using a 5-point frequency rating where after reverse scoring, 1 = never, 2 = yearly, 3 = monthly, 4 = weekly, and 5 = daily. The average frequency rating of 5 items was used in subsequent analyses (BRFSS items 2-6), where higher average scores indicated more frequent consumption of fruits and vegetables. One item about frequency of consuming fruit juices was excluded to improve internal consistency (BRFSS item 1). Reliability and validity have been established.29 Diabetes knowledge was assessed by the Spoken Knowledge in Low Literacy Patients with Diabetes

(SKILLD) scale.30 The 10-item SKILLD assesses knowledge of glucose management, lifestyle modifications, recognition and treatment of hyper- and hypoglycemia, and activities to prevent long-term consequences of the disease. Answers were considered correct if they were acceptable responses.30 Each item score was summed ranging from 0 to 10, with a high score indicating better knowledge about diabetes. Internal consistency using Kuder-Richardson 20 (KR-20) reliability coefficient (KR-20) of the SKILLD was 0.72 in patients with diabetes. Validity has been established.30 KR-20 for the current study was 0.65. Family support was examined using the Diabetes Family Support Behavior Checklist (DFBC-II) (17 items).31 The DFBC-II uses a 5-point Likert scale to assess perceptions of family member support of the person with T2DM in medication taking, glucose testing, exercise, and diet. Positive and negative items are summed separately, and higher scores indicate stronger perceptions of family support.31 Cronbach’s alphas have been reported as 0.71 and 0.64 in Hispanic subjects with diabetes.32 Cronbach’s alphas for the current study were 0.70 for the supportive subscale and 0.64 for the nonsupportive subscale. Diabetes self-efficacy was examined by the 8-item Stanford Self-Efficacy Scale, which measures the confidence of a person with diabetes to manage diet, exercise, knowledge of blood glucose and the illness, and control over diabetes. Scores range from 1 to 10, from no confidence to totally confident; higher scores indicate greater confidence. Internal reliability was 0.85 in Spanishspeaking adults with diabetes (Spanish version) and 0.83 in English-speaking individuals with diabetes.33 Cronbach’s alpha for the current study was 0.87. Diabetes self-management was measured by the Revised Summary of Diabetes Self-Care Activities (SDSCA) (19 items), which assessed self-care activities in diet, exercise, glucose testing, medication, and foot care.34 The mean number of days per week that activities were performed was calculated on a scale of 0 to 7, and a high mean score indicated better diabetes selfmanagement. Internal consistencies using interitem correlations have ranged from 0.47 to 0.90 in adults with diabetes. The validity of the SDSCA scale is established.34 Validity and reliability of the Spanish version of the SDSCA have been reported.35 Health-related quality of life was measured by the Medical Outcomes Study Short Form (SF-12) Health Survey36 (Spanish version). The SF-12 consists of 2 major

Hu et al Downloaded from at Nat. Taichung Univ. of Sci. & Tech. on April 21, 2014

The Diabetes EDUCATOR 52

constructs: physical health and mental health, with 8 health concepts represented by the 12 items that have a Likert scaling for responses. Subscale scores are transformed from normal scaling to a 0 to 100 standardized score, with higher scores indicating more positive health and better quality of life. The SF-12 has demonstrated good validity37 and has been used with multiple genders, ages, and ethnic populations in multiple settings.21 Cronbach’s alphas for the current study were 0.70 for the physical health component and 0.74 for the mental health component. Data collected on family members included demographic data, diabetes knowledge, blood glucose, and blood pressure at baseline and post-intervention. It took approximately 1 hour for data collection at each data collection point. All written instruments were verbally administered in participants’ language of choice to prevent embarrassment about low literacy. Statistical Analyses

Growth curve analyses were used to model change over time from baseline and to estimate mean differences from pre- to post-intervention and from pre-intervention to 1-month follow-up for participants with diabetes. Growth curve modeling allows for individuals’ trajectories of improvement or decline to be modeled in a parsimonious fashion while incorporating all available observations (as opposed to traditional repeated-measures analysis of variance that is based on complete data). It can be accomplished by way of multilevel analysis, where for a longitudinal study, level-1 units are observations at each occasion and level-2 units are participants,38 and is also among a broader class of models known as random-coefficient models.39 Linear splines with knots located at the mean time for the post-intervention patients40 were incorporated where appropriate. Unlike polynomial effects (eg, quadratic, cubic), linear splines model nonlinearities through piecewise linear terms for time trends, allowing for separate tests of trend for each term.40 Models with linear trends for time or linear splines were adequate using the Akaike information criterion.41 Final models specified random intercepts and random slopes for time with an unstructured covariance structure in mixed-effects regressions.40 When normality assumptions were questionable, ranks of outcome values were alternatively modeled for tests of trends over time. For patients, improvements in self-management of diabetes was tested by assessing whether there were significant improvements in SDSCA scores at post-intervention

compared with pre-intervention and also at 1-month follow-up in growth curve modeling. Using similar methods, improvements in physical activity were assessed in modeling IPAQ-based estimates of mean MET-minutes per week and kilocalories per week. Improvements in diet were assessed from similar modeling of BRFSS fruits and vegetables measures. Metabolic measures such as BMI, A1C, total cholesterol/HDL ratio, and systolic and diastolic blood pressure were analyzed similarly by testing for mean improvements at post-intervention and 1-month follow-up in growth curve analyses. Testing for increases in diabetes knowledge was performed again using growth curve modeling of mean SKILLD scores for score increases at post-intervention or 1-month follow-up relative to pre-intervention. Diabetes self-efficacy improvements were likewise evaluated by assessing whether there were mean improvements in Stanford diabetes selfefficacy scores. Increases in family support were testing in growth curve modeling of mean DFBC-II supportive and nonsupportive scores. Increases in health-related quality of life were tested from modeling of mean SF-12 physical component scores and mental component scores. For family members, paired t tests or Wilcoxon signed rank tests were used to examine change from pre- to post-intervention for each of BMI, waist circumference, blood pressure, MET-minutes per week, kilocalories per week, BRFSS fruits and vegetables, and SKILLD scores. All analyses were performed in SAS version 9.2 (SAS Institute, Cary, North Carolina). A 2-sided P-value 10: –0.401 (–0.636 to −0.165) .0016 –0.087 (–0.244 to 0.071) .2763 –0.006 (–0.075 to 0.063) .8590


SDSCA smoke

SDSCA medications


Hu et al Downloaded from at Nat. Taichung Univ. of Sci. & Tech. on April 21, 2014

The Diabetes EDUCATOR 56

Table 1

(continued) Pre-Intervention Estimate (95% CI)

Post-Interventiona Estimate (95% CI)

1-Month Follow-Upb Estimate (95% CI)

Slope (95% CI) P Value

SKILLD score (KR-20 = 0.65)

4.07 (3.32 to 4.82)

9.08 (8.32 to 9.85)

8.62 (8.08 to 9.17)

BRFSS item 1

3.38 (2.81 to 3.95)

3.59 (3.18 to 4.00)

3.68 (3.19 to 4.17)

BRFSS score (mean of reversed items 2-6) (α = 0.60)

3.46 (3.18 to 3.73)

4.06 (3.87 to 4.25)

3.70 (3.44 to 3.96)

SF-12 Physical Component Score (α = 0.70)

61.53 (53.62 to 69.45)

69.42 (63.30 to 75.54)

72.96 (66.78 to 79.15)

SF-12 Mental Component Score (α = 0.74)

58.40 (52.41 to 64.38)

63.98 (59.51 to 68.46)

66.50 (61.45 to 71.55)

≤10 weeks: 0.501 (0.389 to 0.614) 10: –0.604 (–0.900 to −0.307) .0003 0.021 (–0.026 to 0.067) .3734 ≤10 weeks: 0.060 (0.028 to 0.092) .0006 increment >10: –0.140 (–0.227 to −0.054) .0021 0.788 (0.364 to 1.213) .0006 0.559 (0.123 to 0.994) .0134


Abbreviations: BRFSS, Behavioral Risk Factor Surveillance System; DFBC, Diabetes Family Support Behavior Checklist; MET, metabolic equivalent task; SDSCA, Summary of Diabetes Self-Care Activities; SF-12, Medical Outcomes Study Short Form; SKILLD, Spoken Knowledge in Low Literacy Patients with Diabetes. a Mean time from pre- of all patients post-intervention was approximately 10 weeks. b Mean time from pre- of all patients at 1-month follow-up was approximately 14.5 weeks.

Discussion This pilot study examined the efficacy of an 8-week culturally tailored intervention focusing on diabetes self-management for Hispanic adults with diabetes and their family members. The findings indicated that the intervention had positive effects on participants’ (1) systolic blood pressure, (2) diabetes knowledge, (3) diabetes self-efficacy, (4) selfmanagement of general diet, specific diet, blood glucose

testing, and foot care, (5) fruit and vegetable consumption, and (6) both the physical and mental components of healthrelated quality of life. Also, significant changes were found among family members, including improvements in BMI and diabetes knowledge. For participant with diabetes, clinical improvements were found in A1C, waist circumference, LDL, and family support; in family members, clinical improvements were found in systolic blood pressure, kilocalories burned per week in physical activity, and

Volume 40, Number 1, January/February 2014 Downloaded from at Nat. Taichung Univ. of Sci. & Tech. on April 21, 2014

A Family-Based Diabetes Intervention 57

Table 2

Change in Outcomes at Pre- vs Post-Intervention for Family Members Outcome

Mean Difference

Body mass index, kg/m2 Waist circumference, cm Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg Physical activity, MET-min/wk Physical activity, kcal/wk SKILLD (KR-20 = 0.65) BRFSS fruits and vegetables (α = 0.72)

–0.25 –2.61 –3.94 –1.87 3430.4 5219.7 5.89 0.30

95% CI for Mean Difference (–0.47 to −0.04) (–7.65 to 2.43) (–8.38 to 0.51) (–4.77 to 1.02) (–882.3 to 7743.2) (–784.8 to 11,224.1) (4.90 to 6.89) (–0.01 to 0.61)

P Value .0234 .1830 .0804 .1415 .1147 .0860

A family-based diabetes intervention for Hispanic adults and their family members.

The purpose of this quasi-experimental, 1-group longitudinal study is to examine the effects of a family-based intervention program on diabetes self-m...
385KB Sizes 0 Downloads 0 Views